Surge in AC DC Converter

[KavithaR]

Hi,
We have been using Hi link Module (HLK-5M05) in the motor controller application for AC to DC conversion.
Input voltage is taken from the contactor parallelly.
Whenever the contactor is driven, Spike is observed on the output voltage of the HLK-5M05 module which is causing the microcontroller to malfunction (Halt). System will be start working once the controller goes to power on reset

Input Voltage (AC): 90V ~ 264V
Output Voltage (DC): 5V
Output Current: 1A

We have replaced the Hi-Link (HLK-5M05) module with LS05-26B05R3 having phase to phase input voltage with a 4kV isolation and facing the same issue
Input Voltage (AC): 90V ~ 528V
Output Voltage (DC): 5V
Output Current: 1mA
Isolation Voltage: 4kV
Below is the attached datasheet link for LS05-26B05R3 module
https://www.mornsun-power.com/html/pdf/LS05-26B05R3.html
Kindly suggest the possible solutions and debugging tips
Thanks in advance.

 

[KlausST]

Hi,

according your schematic:
* [+5V_HL] is not connected at all
* [PWR_FLAG] is shorted three times
* unclear what [VCC_5V0], [+5V], [+5VA] are used for ... they are all shorted together.

****
Unknown:
* What conductor?
* what "spike" you are talking about. Voltage, frequency, waveform...
* how you measured the "spike"

****

Output Current: 1mA

So your power supply is 1000 times oversized.
Thus it´s very likely it operates in discontinous burst mode. Causing a lot more ripple than at full load. (as specified in the datasheet)

I miss the 10uF || 100nF capacitors at the output as stated in the datasheet. (not necessary for operation, but necessary to get the specified ripple results)

Klaus

 

[Kavitha R]

Hi KlausST,
Thank you for your valuable reply.

according your schematic:
* [+5V_HL] is not connected at all
* [PWR_FLAG] is shorted three times
* unclear what [VCC_5V0], [+5V], [+5VA] are used for ... they are all shorted together.

+5V_HL is a output of AC-DC Converter
+5V_BB is a external power
VCC_5V0 is a controller input supply

We have used jumper for choosing 5V either from AC DC Converter or External power

[PWR_FLAG] is shorted three times

In Kicad, Power pins are labelled as PWR FLAG. It is not shorted

* What conductor?
* what "spike" you are talking about. Voltage, frequency, waveform...
* how you measured the "spike"

It's a voltage spike. Constant 5V DC is not getting in the output of AC DC Converter. We have observed undershoot and overshoot that is causing a micro controller to malfunction (Halt).

So your power supply is 1000 times oversized.
Thus it´s very likely it operates in discontinous burst mode. Causing a lot more ripple than at full load. (as specified in the datasheet)

Sorry, it's 1A not 1mA

 

[KlausST]

Hi,

+5V_HL is a output of AC-DC Converter
+5V_BB is a external power
VCC_5V0 is a controller input supply

We have used jumper for choosing 5V either from AC DC Converter or External power

I´m not sure what information I should get from this.
If there is a jumper then show it. We need to understand what you want to achieve.

In Kicad, Power pins are labelled as PWR FLAG. It is not shorted

I´m not familiar with Kicad. Can´t comprehend what it is used for.

It's a voltage spike. Constant 5V DC is not getting in the output of AC DC Converter. We have observed undershoot and overshoot that is causing a micro controller to malfunction (Halt).

Again: voltage, frequency, waveform! Text is useless!
(it´s a long winding way to my mother. How long will I need? ... )
And: a microcontroller has a datasheet. It tells you the valid supply voltage range. We can´t tell you whether your supply and microcontroller matches.

Sorry, it's 1A not 1mA

Ok. Still I doubt it is perfectly smooth 1.000A. Iff you ant us to analze it you need to give useful informations: Min current, max current, wavefrom...

******
All in all. According my experince a hanging microcontroller is unlikely to be caused by a well woring ACDC convereter output ... AS LONG as you cared about ACDC converter datasheet recommendations, microcontroller datasheet recommendations, bypass capacitors, low impedacne PCB layout with solid GND plane.

And there are power supply monitors that give a clear RESET to the microcontroller in case of power supply problems. It will restatr your microcontroller and prevent from hang.

****
I understand you focus on the /too) tiny schematic you gave in post#1. But I think the problem is else where.
So if you want quick solution to the real problem: show the whole story.

Klaus

 

[KavithaR]

Hi Klaus,

Below attached is the overall block diagram of the system where the power flow is indicated. Kindly suggest what can be done to minimise the surge

 

[D.A.(Tony)Stewart]

You have an EMI ingress problem assumed to be conductive from contactor surge and maybe some ripple under/overshoot (?) not clearly shown or defined.

Once we identify the cause with an equivalent circuit, the solution will be easy. But until you can scope the problem and layout to be transferred into some logical diagram, we can only guess.

Is it capacitive coupling from dV/dt from AC to DC?
Is it inductive coupling from a ground loop, power loop or mutual coupling from dI/dt and wire inductance?
Are the egress wires and ingress DC wires shielded or twisted pair or at right angles?
Is it Common Mode noise ingress that can be isolated with a Balun and Pi Filter ?

etc etc. It is up to you to use judgement and 1st principles or RLC characteristics to isolate this cause by; making it worse or better with changes to see it it mimics inductive or capacitive or both coupling by shielding, filtering and/or CM choke or Faraday cage test. Or it may be purely conductive interference mitigated by your measurements with a Diff Probe or two 10:1 probes twisted and displayed in A-B mode. Then measured for conductive isolation using a Vector Network Analyzer for s21 or adding ferrite beads and film/NP0 caps for low ESR above 20 MHz.

You must give far more test and photo information to analyze in this design review to solve quickly.

Right now , all I see is a big loop antenna around your block diagram. lol

 

[KlausST]

Hi,

You must give far more test and photo information to analyze in this design review to solve quickly.

The OP does not want to provide any information about the "spike" ... I don´t want to annoy him with repeatedly asking for the same informations.
Thus I better leave this thread.

Klaus

 

[D.A.(Tony)Stewart]

Hi,


The OP does not want to provide any information about the "spike" ... I don´t want to annoy him with repeatedly asking for the same informations.
Thus I better leave this thread.

Klaus

We just need patience until he learns what is needed to test & show to define the problem better.

What is EMI? | Keystone Compliance

Electromagnetic Interference (EMI) is a disturbance in an electrical circuit that occurs from another source through electromagnetic induction or radiation.

keystonecompliance.com